Why do I bother?

Just what the name says.
thomh
Posts: 124
Joined: Wed Sep 17, 2003 3:16 am
Location: Norway

Why do I bother?

Postby thomh » Sat Oct 18, 2003 11:08 am

Check this thread:

http://www.stevehoffman.tv/forums/showt ... adid=22998

I find it amusing that even though with all the discussion of digital over there, few (except for Gardo) don't even have a clue about sampling.
Thom

User avatar
Rspaight
Posts: 4386
Joined: Wed Apr 30, 2003 10:48 am
Location: The Reality-Based Community
Contact:

Postby Rspaight » Sat Oct 18, 2003 12:04 pm

Digital samples the music, which by definition means it is taking momentary "snapshots" of the music as it plays. The sampling frequency can be extremely high, but even so there are still instants where no music is sampled, which means that a moment of the music is not stored for reproduction. Now, others can argue whether or not this makes a difference in sound quality (I don't have enough time or energy to), but this means that information is lost.


Oh, fer cryin' out loud.

Does anyone ever take the time to think about why we measure these things in Hz? Which is a fancy way of saying cycles per second? Because sound is a vibration. The faster it vibrates, the higher the tone.

There is no information that is "skipped" by some "snapshot" process. Sound is not some sort of infinitely complex stream that can only be captured by the "magic" of a vinyl record spinning at 33 1/3 rpm -- it is a series of easily measurable events happening at easily measureable rates. Period. As long as you're sampling fast enough to capture all of these events, you've captured the sound.

There's an argument to be made that 44.1 is too low, and some audible effect could be caused by putting the limit at 22.05 kHz. But that's not the fault of digital, it's the fault of a specific implementation of digital.

Similarly, you could argue that 16 bits isn't enough, though I have to wonder about that. It's plenty to capture dynamic range far in excess of any LP. Which makes statements like this:

Never thought it possible for digital to have the dynamic range of Lp's.


especially puzzling.

When someone expresses a preference for some LP over some CD (or vice versa), we're talking about different tape sources and different mastering (and euphonic distortion and placebo effect) here, not any inherent advantage in one medium over another.

Ryan
RQOTW: "I'll make sure that our future is defined not by the letters ACLU, but by the letters USA." -- Mitt Romney

thomh
Posts: 124
Joined: Wed Sep 17, 2003 3:16 am
Location: Norway

Postby thomh » Sat Oct 18, 2003 2:34 pm

Rspaight wrote:Does anyone ever take the time to think about why we measure these things in Hz? Which is a fancy way of saying cycles per second? Because sound is a vibration. The faster it vibrates, the higher the tone.

There is no information that is "skipped" by some "snapshot" process. Sound is not some sort of infinitely complex stream that can only be captured by the "magic" of a vinyl record spinning at 33 1/3 rpm -- it is a series of easily measurable events happening at easily measureable rates. Period. As long as you're sampling fast enough to capture all of these events, you've captured the sound.


Extactly. The misconception that I see all time is regarding sampling is the notion that the numbers 16/44 or 24/96 has something to do with the number of "snapshots" taken. What I have been trying to explain to these people in this thread, and which Gardo nailed, is that within the frequency response limit of either 16/44 or 24/96, there are only 2 samples needed to capture the waveform with *no* missing information whatsoever as there is *only* one waveform which will pass through both points. This does not seem to register with them.

There's an argument to be made that 44.1 is too low, and some audible effect could be caused by putting the limit at 22.05 kHz. But that's not the fault of digital, it's the fault of a specific implementation of digital.


Consider the following: If the input signal has a dynamic range of 90dB and a FR of 20Hz-20kHz then *all* the information is captured by an accurate 16/44 ADC. This should pretty take care of 99.9% of all master tapes. Which is probably the reason I cannot tell the difference between a proper mastered CD and SACD or DVD-A. Comparisons of both Santana and Neil Young comes to mind. There just is not any more information on the tape to benefit the hi-rez formats. However, I do think that 24/96 is beneficial in the recording process due to greater ease of coping with accidental overloads or lots of EQ.

With regards to the dynamic range of LPs being greater than ordinary Redbook, consider that you digitally record an LP and capture all the "magic" in the grooves. What does that tell you about transparency and dynamic range. Try going the other way with a really dynamically recorded CD.
Thom

User avatar
Rspaight
Posts: 4386
Joined: Wed Apr 30, 2003 10:48 am
Location: The Reality-Based Community
Contact:

Postby Rspaight » Sat Oct 18, 2003 3:09 pm

With regards to the dynamic range of LPs being greater than ordinary Redbook, consider that you digitally record an LP and capture all the "magic" in the grooves. What does that tell you about transparency and dynamic range. Try going the other way with a really dynamically recorded CD.


Even Steve Hoffman will tell you that CD lets you more neutrally present the sound on the original master, without the EQ and other tweaking needed to accomodate the physical limits of vinyl.

Ryan
RQOTW: "I'll make sure that our future is defined not by the letters ACLU, but by the letters USA." -- Mitt Romney

User avatar
lukpac
Top Dog and Sellout
Posts: 4591
Joined: Wed Apr 02, 2003 11:51 pm
Location: Madison, WI
Contact:

Postby lukpac » Sat Oct 18, 2003 4:08 pm

thomh wrote:Extactly. The misconception that I see all time is regarding sampling is the notion that the numbers 16/44 or 24/96 has something to do with the number of "snapshots" taken.


Well, they do, actually. 44 means 44100 individual "snapshots" per second. 16 of course means there are 65536 possible values for each of those snapshots.

What I have been trying to explain to these people in this thread, and which Gardo nailed, is that within the frequency response limit of either 16/44 or 24/96, there are only 2 samples needed to capture the waveform with *no* missing information whatsoever as there is *only* one waveform which will pass through both points. This does not seem to register with them.


Not quite. You can obviously plot an infinite number of waves through two points. Irrelevant, but true nevertheless. However, it *is* technically possible to have different waves end up the same digitally. They would just have to be closer than the 16-bit precision allows.

Even after clearing all that up, though, I honestly think 16/44 is fine. I can't hear any difference between my LPs and the CD-Rs I make from them, although of course I'm using inferior equipment...
"I know because it is impossible for a tape to hold the compression levels of these treble boosted MFSL's like Something/Anything. The metal particulate on the tape would shatter and all you'd hear is distortion if even that." - VD

thomh
Posts: 124
Joined: Wed Sep 17, 2003 3:16 am
Location: Norway

Postby thomh » Sat Oct 18, 2003 4:17 pm

Thanks Luke. Noted and logged.
Thom

thomh
Posts: 124
Joined: Wed Sep 17, 2003 3:16 am
Location: Norway

Postby thomh » Sat Oct 18, 2003 4:20 pm

Rspaight wrote:
With regards to the dynamic range of LPs being greater than ordinary Redbook, consider that you digitally record an LP and capture all the "magic" in the grooves. What does that tell you about transparency and dynamic range. Try going the other way with a really dynamically recorded CD.


Even Steve Hoffman will tell you that CD lets you more neutrally present the sound on the original master, without the EQ and other tweaking needed to accomodate the physical limits of vinyl.

Ryan


Here are a couple of interesting quotes from Steve taken from this interview in 1997.

http://home.earthlink.net/~mercmoon/hoffmanint.htm:

DMG - Maybe, though with higher resolution formats there might be more detail...

Steve - You know, when you say higher resolution, I get scared. To me, higher resolution means Krell. It's like, how much is there anyway? Lets say the Lourve museum loans you the Mona Lisa for a couple of days. Ok, so here's the Mona Lisa. Where do you show it? Do you hang it up in the house with a nice, soft light? Or do you take it outside with the sun shining on it, where all you're going to see are the scratches and cracks. How much resolution is there? The less resolution there is on that, the better its going to look. That's my only fear. When you look at a piece of color printing, and all you see are colored dots all over the page, you wonder how much resolution was there to begin with. When you listen to the Beatles' I Want To Hold Your Hand on a really good system, it's
disappointing. It's not a sonic masterpiece. It's a good song, but not a great recording. So, words like 'higher resolution' to me, scare me a little bit.

DMG - Most of the audio bandwith can be handled very well by 16bit CD.

Steve - Yes.


DMG - So, is 20 or 24 bit going to get to a level where the emotion is completely gone through more imperfections being brought out?

Steve - That's exactly why higher resolution scares me. With higher res on that recording, you're going to hear more high end that shouldn't be there. People would be surprised if they had a meter hooked up to their
system and could see that on 90% of their records and discs, above 9500K, there's nothing there. What you're hearing up there is tape hiss, or some other thing that isn't part of the music. When you extend everything up to that range, there's nothing up there but noise.

DMG - Well, it seems people don't want to admit that their hearing basically falls off dramatically around 13 or 14k.

Steve - Not only that, but the famous Neumann mikes have a giant peak around 6k, and from there it's all downhill. You know, you have to examine what it is you want from music, and if you're one of those people
that relies on charts or instrument readings, and you say, 'Well, my system reproduces from 5 cycles to 9 million cycles'. Well, that's great, but you sure aren't hearing it. Numbers aren't everything. You have to trust your ears, and I always tell people that.


And something he told me when I asked him about his mastering of In The Ghetto:

I transferred IN THE GHETTO flat from the two-track master mix. The DCC version is how the actual tape sounds.


Lately however, when has gotten involved with SACD, it seems he's changed his tune a little bit for whatever reason.
Thom

thomh
Posts: 124
Joined: Wed Sep 17, 2003 3:16 am
Location: Norway

Postby thomh » Sat Oct 18, 2003 4:55 pm

In this thread, which I see Luke participated, the master tapes which he has spent close to 20 years listening to and digitizing has suddenly seen a rise in dynamic range.

http://www.stevehoffman.tv/forums/showthread.php?s=&threadid=12530&highlight=Whats+the+Worst+thing+about+SACDs

GoldenGuy,

The boundaries of Digital PCM have already been pushed to their limit if I play back a master tape vs. the PCM copy and hear things like echo fall off on the digital. If they sounded the same in an A/B all our problems would be over and my job would be so much easier.


And here he snaps at Pepzhez for really just putting forth some of the same thoughts that Steve had in the 1997 interview:

http://www.stevehoffman.tv/forums/showthread.php?s=&threadid=19886&highlight=sacd

Read into it what you will.....
Thom

thomh
Posts: 124
Joined: Wed Sep 17, 2003 3:16 am
Location: Norway

Postby thomh » Sat Oct 18, 2003 5:18 pm

lukpac wrote:Even after clearing all that up, though, I honestly think 16/44 is fine. I can't hear any difference between my LPs and the CD-Rs I make from them, although of course I'm using inferior equipment...


:lol:

Yeah, that is what I have been told as well. It seems inorder to appreciate the "finer details" of SACD, as with vinyl, you need to invest in them *really* big shiny toys.

If that is what it takes, how in the hell is Sony gonna market this to Joe Blow. Seems pretty much DOA to me. Unless the marketing people have something up their sleeve.....
Thom

User avatar
krabapple
Posts: 1615
Joined: Wed Apr 16, 2003 4:19 pm

Re: Why do I bother?

Postby krabapple » Tue Oct 21, 2003 5:49 pm

thomh wrote:Check this thread:

http://www.stevehoffman.tv/forums/showt ... adid=22998

I find it amusing that even though with all the discussion of digital over there, few (except for Gardo) don't even have a clue about sampling.


Gary's avatars always creep me out.

Basically you're arguing with know-nothings; they'll never be convinced by stupid FACTS.
THey'll just keep repeating their misconceptions, over and over and over.

They're like the anti-evolutionists who always say, "If man evolved from monkeys HOW COME THERE ARE STILL MONKEYS TODAY HUH??" THe whole POV is just invinciibly loaded with mistakes and incomprehension of fact. You have to correct multiple errors in just one sentence, and then try to move on from there. It's hardly worth the effort sometimes.

thomh
Posts: 124
Joined: Wed Sep 17, 2003 3:16 am
Location: Norway

Postby thomh » Wed Oct 22, 2003 2:33 am

Here is another fun read regarding interconnect and cable "sound".

http://www.stevehoffman.tv/forums/showt ... genumber=1

As I argued, this should be easy enough to prove. But as usual, the people making extraordinary claims, fail to step up to the plate and instead resorts to things like questioning the test conditions.

And Steve introduces a spankin' new "best there is" product from his buddy Grover.
Thom

User avatar
Rspaight
Posts: 4386
Joined: Wed Apr 30, 2003 10:48 am
Location: The Reality-Based Community
Contact:

Postby Rspaight » Wed Oct 22, 2003 8:16 am

Dave on posting references to newsgroups:

To create a link...Highlight the address in the address bar while viewing the page, right-click and save it. Come here to post and right-click and then paste in your post. Viola! A link.


Stop. My sides are hurting.

Viola!

Ryan
RQOTW: "I'll make sure that our future is defined not by the letters ACLU, but by the letters USA." -- Mitt Romney

User avatar
Rspaight
Posts: 4386
Joined: Wed Apr 30, 2003 10:48 am
Location: The Reality-Based Community
Contact:

Postby Rspaight » Mon Nov 03, 2003 5:27 pm

RQOTW: "I'll make sure that our future is defined not by the letters ACLU, but by the letters USA." -- Mitt Romney

User avatar
Xenu
Sellout
Posts: 2209
Joined: Mon Apr 07, 2003 8:15 pm

Postby Xenu » Mon Nov 03, 2003 6:40 pm

FWIW, there's finally progress being made on the whole UDI vs UDII thread, with a few people (thanks, Ryan!) standing firm that when digital says that two discs are identical, barring jitter, etc., they *are* identical.

Dave's using the "But Steve promised my ears are more accurate" approach.

thomh
Posts: 124
Joined: Wed Sep 17, 2003 3:16 am
Location: Norway

Postby thomh » Mon Nov 03, 2003 6:53 pm

Rspaight wrote:The hilarity continues:

http://www.stevehoffman.tv/forums/showt ... adid=23790

Ryan


Oh, my....I would love to get involved in that thread and respond to bunk like this:

Analog = continuous
Digital = sampled, algorithms fill in the blanks.

Technical measurements tell me the later is more accurate - yet I still haunt garage sales to purchase used vinyl for the former. A bit of crackle is preferable to me then the "blanks" I shouldn't notice.

Guess I need some digital ears.


Jeez......
Thom